fast computation of eigenvalues of companion comrade and
play

Fast computation of eigenvalues of companion, comrade, and related - PowerPoint PPT Presentation

Apr 08, 2024 •32 likes •1.4k views

Fast computation of eigenvalues of companion, comrade, and related matrices David S. Watkins Department of Mathematics Washington State University Providence, June 2013 David S. Watkins companion, comrade, and related matrices Collaborators

  1. How do we compute the eigenvalues? narrow band plus spike  × × × × × ×  × × ×     × × ×   A =   × × ×     × × ×   × × Francis’s bulge-chasing algorithm (“implicitly-shifted QR ”) O ( n 3 ) time, O ( n 2 ) space Is there a more economical approach? David S. Watkins companion, comrade, and related matrices

  2. How do we compute the eigenvalues? narrow band plus spike  × × × × × ×  × × ×     × × ×   A =   × × ×     × × ×   × × Francis’s bulge-chasing algorithm (“implicitly-shifted QR ”) O ( n 3 ) time, O ( n 2 ) space Is there a more economical approach? David S. Watkins companion, comrade, and related matrices

  3. How do we compute the eigenvalues? narrow band plus spike  × × × × × ×  × × ×     × × ×   A =   × × ×     × × ×   × × Francis’s bulge-chasing algorithm (“implicitly-shifted QR ”) O ( n 3 ) time, O ( n 2 ) space Is there a more economical approach? David S. Watkins companion, comrade, and related matrices

  4. How do we compute the eigenvalues? narrow band plus spike  × × × × × ×  × × ×     × × ×   A =   × × ×     × × ×   × × Francis’s bulge-chasing algorithm (“implicitly-shifted QR ”) O ( n 3 ) time, O ( n 2 ) space Is there a more economical approach? David S. Watkins companion, comrade, and related matrices

  5. How do we compute the eigenvalues? banded plus spike  × × × × × ×  × × ×     × × ×   A =   × × ×     × × ×   × × Exploit rank structure (generators) We pursue a different approach. David S. Watkins companion, comrade, and related matrices

  6. How do we compute the eigenvalues? banded plus spike  × × × × × ×  × × ×     × × ×   A =   × × ×     × × ×   × × Exploit rank structure (generators) We pursue a different approach. David S. Watkins companion, comrade, and related matrices

  7. How do we compute the eigenvalues? banded plus spike  × × × × × ×  × × ×     × × ×   A =   × × ×     × × ×   × × Exploit rank structure (generators) We pursue a different approach. David S. Watkins companion, comrade, and related matrices

  8. Fast Methods We have developed great methods that are . . . lightning fast unstable David S. Watkins companion, comrade, and related matrices

  9. Fast Methods We have developed great methods that are . . . lightning fast unstable David S. Watkins companion, comrade, and related matrices

  10. Fast Methods We have developed great methods that are . . . lightning fast unstable David S. Watkins companion, comrade, and related matrices

  11. A useful factorization (by Gaussian elimination) × × × × × ×   × ×   × ×     × ×   × ×   × × David S. Watkins companion, comrade, and related matrices

  12. A useful factorization (by Gaussian elimination) × × × × × × × ×     � � × × × × × × × ×     × × × ×     =     × × × ×     × × × ×     × × × × David S. Watkins companion, comrade, and related matrices

  13. A useful factorization (by Gaussian elimination) × × × × × × × ×     � � × × 0 × × × × ×     × × × ×     =     × × × ×     × × × ×     × × × × David S. Watkins companion, comrade, and related matrices

  14. A useful factorization (by Gaussian elimination) × × × × × × × ×     � � × × × × �     � × × × × × × ×     =     × × × ×     × × × ×     × × × × David S. Watkins companion, comrade, and related matrices

  15. A useful factorization (by Gaussian elimination) × × × × × × × ×     � � × × × × �     � × × 0 × × × ×     =     × × × ×     × × × ×     × × × × David S. Watkins companion, comrade, and related matrices

  16. A useful factorization (by Gaussian elimination) × × × × × × × ×     � � × × × × �     � × × × ×     � =     � × × × × × ×     × × × ×     × × × × David S. Watkins companion, comrade, and related matrices

  17. A useful factorization (by Gaussian elimination) × × × × × × × ×     � � × × × × �     � × × × ×     � =     � × × 0 × × ×     × × × ×     × × × × David S. Watkins companion, comrade, and related matrices

  18. A useful factorization (by Gaussian elimination)  × × × × × ×   × ×  � � × × × × �     � × × × × �     =     � × × × × ×      × ×   × ×  × × × × and so on . . . to (banded) triangular form. David S. Watkins companion, comrade, and related matrices

  19. A useful factorization (by Gaussian elimination) We get the factorization × ×   � � × × �   � × × �   A =   � × × �   �  × ×  � � × core transformations times banded upper triangular band width = length of recurrence David S. Watkins companion, comrade, and related matrices

  20. What core transformations look like       × × 1 1 � � × × × × 1 � =       � 1 × × × ×       � � 1 1 × × David S. Watkins companion, comrade, and related matrices

  21. Bulge Chasing Algorithm  × ×  � � × × �   � × × �   �   × × �   �  × ×  � � × We will develop a nonunitary bulge-chasing algorithm that preserves this factorization. Note: Storage is O ( n ) (about 6 n for this case). We have single-shift and double-shift variants. David S. Watkins companion, comrade, and related matrices

  22. Bulge Chasing Algorithm  × ×  � � × × �   � × × �   �   × × �   �  × ×  � � × We will develop a nonunitary bulge-chasing algorithm that preserves this factorization. Note: Storage is O ( n ) (about 6 n for this case). We have single-shift and double-shift variants. David S. Watkins companion, comrade, and related matrices

  23. Bulge Chasing Algorithm  × ×  � � × × �   � × × �   �   × × �   �  × ×  � � × We will develop a nonunitary bulge-chasing algorithm that preserves this factorization. Note: Storage is O ( n ) (about 6 n for this case). We have single-shift and double-shift variants. David S. Watkins companion, comrade, and related matrices

  24. Bulge Chasing Algorithm  × ×  � � × × �   � × × �   �   × × �   �  × ×  � � × We will develop a nonunitary bulge-chasing algorithm that preserves this factorization. Note: Storage is O ( n ) (about 6 n for this case). We have single-shift and double-shift variants. David S. Watkins companion, comrade, and related matrices

  25. Bulge Chasing Algorithm (single-shift variant) × ×   � � × × �   � × × �     � × × �   �  × ×  � � × David S. Watkins companion, comrade, and related matrices

  26. Bulge Chasing Algorithm (single-shift variant) × ×   � � � � � � × × �   � × × �     � × × �   �  × ×  � � × David S. Watkins companion, comrade, and related matrices

  27. Bulge Chasing Algorithm (single-shift variant) × ×   � � � � × × �   � × × �     � × × �   �  × ×  � � × David S. Watkins companion, comrade, and related matrices

  28. Passing Gauss transform through triangular matrix Blue core transformations are Gauss transforms. � × � × � × × � × � × � � = � = � × × × × × � × × × × × We do this at every opportunity. From this point on, we suppress the triangular matrix. David S. Watkins companion, comrade, and related matrices

  29. Passing Gauss transform through triangular matrix Blue core transformations are Gauss transforms. � × � × � × × � × � × � � = � = � × × × × × � × × × × × We do this at every opportunity. From this point on, we suppress the triangular matrix. David S. Watkins companion, comrade, and related matrices

  30. Passing Gauss transform through triangular matrix Blue core transformations are Gauss transforms. � × � × � × × � × � × � � = � = � × × × × × � × × × × × We do this at every opportunity. From this point on, we suppress the triangular matrix. David S. Watkins companion, comrade, and related matrices

  31. Bulge Chasing Algorithm (single-shift variant) × ×   � � � � × × �   � × × �     � × × �   �  × ×  � � × David S. Watkins companion, comrade, and related matrices

  32. Bulge Chasing Algorithm (single-shift variant) × ×   � � � � × × �   � × × �     � × × �   �  × ×  � � × David S. Watkins companion, comrade, and related matrices

  33. Bulge Chasing Algorithm (single-shift variant) � � � � � � � � � � � � David S. Watkins companion, comrade, and related matrices

  34. Bulge Chasing Algorithm (single-shift variant) � � � � � � � � � � � � David S. Watkins companion, comrade, and related matrices

  35. Bulge Chasing Algorithm (single-shift variant) � � � � � � � � � � � � David S. Watkins companion, comrade, and related matrices

  36. Bulge Chasing Algorithm (single-shift variant) � � � � � � � � � � � � David S. Watkins companion, comrade, and related matrices

  37. Turnover operation Can we do this? � � � � � � ⇔ � � � � � � crucial question Pessimistic answer: No! (not always) Optimistic answer: Yes! (almost always) David S. Watkins companion, comrade, and related matrices

  38. Turnover operation Can we do this? � � � � � � ⇔ � � � � � � crucial question Pessimistic answer: No! (not always) Optimistic answer: Yes! (almost always) David S. Watkins companion, comrade, and related matrices

  39. Turnover operation Can we do this? � � � � � � ⇔ � � � � � � crucial question Pessimistic answer: No! (not always) Optimistic answer: Yes! (almost always) David S. Watkins companion, comrade, and related matrices

  40. Turnover operation Can we do this? � � � � � � ⇔ � � � � � � crucial question Pessimistic answer: No! (not always) Optimistic answer: Yes! (almost always) David S. Watkins companion, comrade, and related matrices

  41. Turnover operation � × � × × × � × × � � � � � � � = = × × × × × × � � 0 × × × × × Red submatrix still has rank one. David S. Watkins companion, comrade, and related matrices

  42. Turnover operation � × � × × × � × × � � � � � � � = = × × × × × × � � 0 × × × × × Red submatrix still has rank one. David S. Watkins companion, comrade, and related matrices

  43. Turnover operation So . . . � × � × × × � × × � � = = � × × × × × × � � � � � � × × × 0 × × Gauss transform does not disturb 2 × 2 submatrix. David S. Watkins companion, comrade, and related matrices

  44. Turnover operation So . . . � × � × × × � × × � � = = � × × × × × × � � � � � � × × × 0 × × Gauss transform does not disturb 2 × 2 submatrix. David S. Watkins companion, comrade, and related matrices

  45. Turnover operation So . . . � × � × × × � × × � � = = � × × × × × × � � � � � � × × × 0 × × Gauss transform does not disturb 2 × 2 submatrix. David S. Watkins companion, comrade, and related matrices

  46. Turnover operation Complete turnover looks like this: � � � � � = � � � � � � � David S. Watkins companion, comrade, and related matrices

  47. Bulge Chasing Algorithm (single-shift variant) � � � � � � � � � � � � David S. Watkins companion, comrade, and related matrices

  48. Bulge Chasing Algorithm (single-shift variant) � � � � � � � � � � � � David S. Watkins companion, comrade, and related matrices

  49. Bulge Chasing Algorithm (single-shift variant) � � � � � � � � � � � � David S. Watkins companion, comrade, and related matrices

  50. Bulge Chasing Algorithm (single-shift variant) � � � � � � � � � � � � David S. Watkins companion, comrade, and related matrices

  51. Bulge Chasing Algorithm (single-shift variant) � � � � � � � � � � � � David S. Watkins companion, comrade, and related matrices

  52. Bulge Chasing Algorithm (single-shift variant) � � � � � � � � � � � � David S. Watkins companion, comrade, and related matrices

  53. Bulge Chasing Algorithm (single-shift variant) � � � � � � � � � � � � David S. Watkins companion, comrade, and related matrices

  54. Bulge Chasing Algorithm (single-shift variant) � � � � � � � � � � � � David S. Watkins companion, comrade, and related matrices

  55. Bulge Chasing Algorithm (single-shift variant) � � � � � � � � � � � � David S. Watkins companion, comrade, and related matrices

  56. Bulge Chasing Algorithm (single-shift variant) � � � � � � � � � � � � David S. Watkins companion, comrade, and related matrices

  57. Bulge Chasing Algorithm (single-shift variant) � � � � � � � � � � � � David S. Watkins companion, comrade, and related matrices

  58. Bulge Chasing Algorithm (single-shift variant) � � � � � � � � � � David S. Watkins companion, comrade, and related matrices

  59. Bulge Chasing Algorithm (single-shift variant) � � � � � � � � � � Done! Now repeat again, again, and again David S. Watkins companion, comrade, and related matrices

  60. Bulge Chasing Algorithm (single-shift variant) � � � � � � � � � � Done! Now repeat again, again, and again David S. Watkins companion, comrade, and related matrices

  61. Bulge Chasing Algorithm (single-shift variant) � � � � � � � � � � Done! Now repeat again, again, and again David S. Watkins companion, comrade, and related matrices

  62. Bulge Chasing Algorithm (single-shift variant) � � � � � � � � � � Done! Now repeat again, again, and again David S. Watkins companion, comrade, and related matrices

  63. Bulge Chasing Algorithm (single-shift variant) � � � � � � � � � � Done! Now repeat again, again, and again David S. Watkins companion, comrade, and related matrices

  64. Computational complexity O ( n ) storage (everything in cache) O ( n ) flops per iteration O ( n ) iterations O ( n 2 ) flops total David S. Watkins companion, comrade, and related matrices

  65. Computational complexity O ( n ) storage (everything in cache) O ( n ) flops per iteration O ( n ) iterations O ( n 2 ) flops total David S. Watkins companion, comrade, and related matrices

  66. Computational complexity O ( n ) storage (everything in cache) O ( n ) flops per iteration O ( n ) iterations O ( n 2 ) flops total David S. Watkins companion, comrade, and related matrices

  67. Computational complexity O ( n ) storage (everything in cache) O ( n ) flops per iteration O ( n ) iterations O ( n 2 ) flops total David S. Watkins companion, comrade, and related matrices

  68. Computational complexity O ( n ) storage (everything in cache) O ( n ) flops per iteration O ( n ) iterations O ( n 2 ) flops total David S. Watkins companion, comrade, and related matrices

  69. Stability issue This is a structured, implicitly-shifted LR algorithm. Unstable! Can we trust the results? No, but we can do a posteriori tests. � ( λ I − A ) v � = | p ( λ ) | | p ( λ ) p ′ ( λ ) | is estimate of error. Newton correction Cost: O ( n 2 ) flops for n roots. David S. Watkins companion, comrade, and related matrices

  70. Stability issue This is a structured, implicitly-shifted LR algorithm. Unstable! Can we trust the results? No, but we can do a posteriori tests. � ( λ I − A ) v � = | p ( λ ) | | p ( λ ) p ′ ( λ ) | is estimate of error. Newton correction Cost: O ( n 2 ) flops for n roots. David S. Watkins companion, comrade, and related matrices

  71. Stability issue This is a structured, implicitly-shifted LR algorithm. Unstable! Can we trust the results? No, but we can do a posteriori tests. � ( λ I − A ) v � = | p ( λ ) | | p ( λ ) p ′ ( λ ) | is estimate of error. Newton correction Cost: O ( n 2 ) flops for n roots. David S. Watkins companion, comrade, and related matrices

  72. Stability issue This is a structured, implicitly-shifted LR algorithm. Unstable! Can we trust the results? No, but we can do a posteriori tests. � ( λ I − A ) v � = | p ( λ ) | | p ( λ ) p ′ ( λ ) | is estimate of error. Newton correction Cost: O ( n 2 ) flops for n roots. David S. Watkins companion, comrade, and related matrices

  73. Stability issue This is a structured, implicitly-shifted LR algorithm. Unstable! Can we trust the results? No, but we can do a posteriori tests. � ( λ I − A ) v � = | p ( λ ) | | p ( λ ) p ′ ( λ ) | is estimate of error. Newton correction Cost: O ( n 2 ) flops for n roots. David S. Watkins companion, comrade, and related matrices

  74. Stability issue This is a structured, implicitly-shifted LR algorithm. Unstable! Can we trust the results? No, but we can do a posteriori tests. � ( λ I − A ) v � = | p ( λ ) | | p ( λ ) p ′ ( λ ) | is estimate of error. Newton correction Cost: O ( n 2 ) flops for n roots. David S. Watkins companion, comrade, and related matrices

  75. Stability issue This is a structured, implicitly-shifted LR algorithm. Unstable! Can we trust the results? No, but we can do a posteriori tests. � ( λ I − A ) v � = | p ( λ ) | | p ( λ ) p ′ ( λ ) | is estimate of error. Newton correction Cost: O ( n 2 ) flops for n roots. David S. Watkins companion, comrade, and related matrices

  76. Stability issue This is a structured, implicitly-shifted LR algorithm. Unstable! Can we trust the results? No, but we can do a posteriori tests. � ( λ I − A ) v � = | p ( λ ) | | p ( λ ) p ′ ( λ ) | is estimate of error. Newton correction Cost: O ( n 2 ) flops for n roots. David S. Watkins companion, comrade, and related matrices

  77. Stability issue This is a structured, implicitly-shifted LR algorithm. Unstable! Can we trust the results? No, but we can do a posteriori tests. � ( λ I − A ) v � = | p ( λ ) | | p ( λ ) p ′ ( λ ) | is estimate of error. Newton correction Cost: O ( n 2 ) flops for n roots. David S. Watkins companion, comrade, and related matrices

  78. Stability issue This is a structured, implicitly-shifted LR algorithm. Unstable! Can we trust the results? No, but we can do a posteriori tests. � ( λ I − A ) v � = | p ( λ ) | | p ( λ ) p ′ ( λ ) | is estimate of error. Newton correction Cost: O ( n 2 ) flops for n roots. David S. Watkins companion, comrade, and related matrices

  79. Performance First example: companion matrices, complex, single-shift code Contestants ( O ( n 3 )) LAPACK code ZHSEQR BBEGG (quasiseparable, unitary) AVW our code David S. Watkins companion, comrade, and related matrices

Recommend

Fast Stable Computation of the Eigenvalues of Companion Matrices David S. Watkins Department of

Fast Stable Computation of the Eigenvalues of Companion Matrices David S. Watkins Department of

Fast Stable Computation of the Eigenvalues of Companion Matrices David S. Watkins Department of Mathematics Washington State University Spa, Belgium, 2014 David S. Watkins Eigenvalues of Companion Matrices Collaborators This is joint work

1.57k views • 111 slides
Computing the eigenvalues of a Joint work with Raf Vandebril and Paul Van Dooren companion

Computing the eigenvalues of a Joint work with Raf Vandebril and Paul Van Dooren companion

Computing the eigenvalues of a companion matrix Marc Van Barel Computing the eigenvalues of a Joint work with Raf Vandebril and Paul Van Dooren companion matrix Structured Linear Algebra Problems: Analysis, Algorithms, and Applications

1.12k views • 64 slides
INTRO COMRADE -1992- Community Outreach for Orang Asli Development COMRADE is one of the

INTRO COMRADE -1992- Community Outreach for Orang Asli Development COMRADE is one of the

DUCTION INTRO COMRADE -1992- Community Outreach for Orang Asli Development COMRADE is one of the leading clubs in IIUM that focus on the development of Orang Asli. Towards achieving the mission and vision of B- OSEM project- Be With

433 views • 19 slides
( A I ) = 0 However, there is no exact formula for polynomials of degree five

( A I ) = 0 However, there is no exact formula for polynomials of degree five

Lesson 21 C OMPUTATION OF EIGENVALUES We consider the problem of computation of eigenvalues To make life (much!) simpler, let's assume we have a symmetric matrix A R n n , which has decomposition A = Q Q Unlike previous

503 views • 26 slides
Chapter 12: Iterative Methods ES 240: Scientific and Engineering Computation. Iterative Methods

Chapter 12: Iterative Methods ES 240: Scientific and Engineering Computation. Iterative Methods

ES 240: Scientific and Engineering Computation. Iterative Methods and Eigenvalues Uchechukwu Ofoegbu Temple University Chapter 12: Iterative Methods ES 240: Scientific and Engineering Computation. Iterative Methods and Eigenvalues Gauss-

261 views • 13 slides
OPERATOR SPLITTING METHODS FOR COMPUTATION OF EIGENVALUES OF REGULAR STURM-LIOUVILLE PROBLEMS

OPERATOR SPLITTING METHODS FOR COMPUTATION OF EIGENVALUES OF REGULAR STURM-LIOUVILLE PROBLEMS

OPERATOR SPLITTING METHODS FOR COMPUTATION OF EIGENVALUES OF REGULAR STURM-LIOUVILLE PROBLEMS Ismail G UZEL ismailgzel@gmail.com Dokuz Eyl ul University IZM IR 13/06/2016 Liouville J. Sturm J.C.F OPERATOR SPLITTING METHOD

607 views • 59 slides
Companion Planting: Plants Need Friends, Too! What is Companion Planting? Planting two or

Companion Planting: Plants Need Friends, Too! What is Companion Planting? Planting two or

Companion Planting: Plants Need Friends, Too! What is Companion Planting? Planting two or more plant species in close proximity to culturally benefit each other by attracting or deterring pests, keeping the soil healthy, and improving

858 views • 22 slides
Graph Computation on Computer Cluster? Steep learning curve Cost Overkill for smaller

Graph Computation on Computer Cluster? Steep learning curve Cost Overkill for smaller

MMap Fast Billion-Scale Graph Computation on a PC via Memory Mapping Lead by Zhiyuan (Jerry) Lin Georgia Tech CS Undergrad Now: Stanford 1st year PhD student MMap: Fast Billion-Scale Graph Computation on a PC via Memory Mapping .

130 views • 11 slides
Fast Constant-Time GCD Computation and Modular Inversion Daniel J. Bernstein 1,2 Bo-Yin Yang 3 1

Fast Constant-Time GCD Computation and Modular Inversion Daniel J. Bernstein 1,2 Bo-Yin Yang 3 1

Fast Constant-Time GCD Computation and Modular Inversion Daniel J. Bernstein 1,2 Bo-Yin Yang 3 1 University of Illinois at Chicago 2 Ruhr Universt at Bochum 3 Academia Sinica Monday, August 26, 2019 Summary: Fast, Safe GCD and Inversions

431 views • 23 slides
Sol Fast Distributed Computation Over Slow Networks Fan Lai , Jie You, Xiangfeng Zhu Harsha V.

Sol Fast Distributed Computation Over Slow Networks Fan Lai , Jie You, Xiangfeng Zhu Harsha V.

Sol Fast Distributed Computation Over Slow Networks Fan Lai , Jie You, Xiangfeng Zhu Harsha V. Madhyastha, Mosharaf Chowdhury 1 Distributed Data Processing is Ubiquitous Distributed computation in Local-Area Networks (LAN) To accelerate

1.92k views • 77 slides
Fast computation of power series solutions of systems of differential equations Alin Bostan

Fast computation of power series solutions of systems of differential equations Alin Bostan

Fast computation of power series solutions of systems of differential equations Alin Bostan ALGO, INRIA Rocquencourt joint work with F. Chyzak , F. Ollivier , B. Salvy , E. Schost , A. Sedoglavic Context find fast algorithms for solving (in

911 views • 22 slides
Chapter 5 Eigenvalues and Eigenvectors Section 5.1 Eigenvectors and Eigenvalues Motivation:

Chapter 5 Eigenvalues and Eigenvectors Section 5.1 Eigenvectors and Eigenvalues Motivation:

Chapter 5 Eigenvalues and Eigenvectors Section 5.1 Eigenvectors and Eigenvalues Motivation: Difference equations A Biology Question How to predict a population of rabbits with given dynamics : 1. half of the newborn rabbits survive their first

353 views • 22 slides
Companion House March 2014 Making a referral to Companion House This presentation is designed

Companion House March 2014 Making a referral to Companion House This presentation is designed

Making referrals to Companion House March 2014 Making a referral to Companion House This presentation is designed to provide information to service providers about how to refer a client to Companion House Aims of the presentation To support

750 views • 36 slides
Eigenvalues, Eigenvectors, Matrix Factoring, and Principal Components The eigenvalues and

Eigenvalues, Eigenvectors, Matrix Factoring, and Principal Components The eigenvalues and

Eigenvalues, Eigenvectors, Matrix Factoring, and Principal Components The eigenvalues and eigenvectors of a square matrix play a key role in some important operations in statistics. In particular, they are intimately connected with the

376 views • 25 slides
Section 5.5 Complex Eigenvalues (Part II) Motivation: Complex Versus Two Real Eigenvalues

Section 5.5 Complex Eigenvalues (Part II) Motivation: Complex Versus Two Real Eigenvalues

Section 5.5 Complex Eigenvalues (Part II) Motivation: Complex Versus Two Real Eigenvalues Todays decomposition is very analogous to diagonalization. Theorem Let A be a 2 2 matrix with linearly independent eigenvectors v 1 , v 2 and

700 views • 16 slides
Humanoid Companion Embedded GPUs can make your robotic companion more alive Hi Alex!

Humanoid Companion Embedded GPUs can make your robotic companion more alive Hi Alex!

Smarter Humanoid Companion Embedded GPUs can make your robotic companion more alive Hi Alex! Alexandre Mazel Innovation Software Director Oct 2017 Agenda Overview Innovation Team Presentation Mummer Research Project

399 views • 29 slides
For matrix A ( p p ) with real eigenvalues, define F A , the empirical distribution function of

For matrix A ( p p ) with real eigenvalues, define F A , the empirical distribution function of

For matrix A ( p p ) with real eigenvalues, define F A , the empirical distribution function of the eigenvalues of A , to be F A ( x ) (1 /p ) (number of eigenvalues of A x ) . For and p.d.f. G the Stieltjes transform of G is defined

447 views • 19 slides
For matrix A ( p p ) with real eigenvalues, define F A , the empirical distribution function of

For matrix A ( p p ) with real eigenvalues, define F A , the empirical distribution function of

For matrix A ( p p ) with real eigenvalues, define F A , the empirical distribution function of the eigenvalues of A , to be F A ( x ) (1 /p ) (number of eigenvalues of A x ) . For and p.d.f. G the Stieltjes transform of G is defined

451 views • 24 slides
Eigenvalues, Eigenvectors, and Diagonalization Diagonalization Math 240 Calculus III Summer

Eigenvalues, Eigenvectors, and Diagonalization Diagonalization Math 240 Calculus III Summer

Eigenvalues, Eigenvectors, and Diagonal- ization Math 240 Eigenvalues and Eigenvectors Eigenvalues, Eigenvectors, and Diagonalization Diagonalization Math 240 Calculus III Summer 2015, Session II Thursday, July 16, 2015 Eigenvalues,

171 views • 15 slides
Fast computation of the N th term of an algebraic series in positive characteristic Philippe

Fast computation of the N th term of an algebraic series in positive characteristic Philippe

Fast computation of the N th term of an algebraic series in positive characteristic Philippe Dumas JNCF 2015, Cluny Joint work with Alin Bostan and Gilles Christol Introduction Special case of rational series Algorithms with low complexity

890 views • 65 slides
The power of How fast is sorting? parallel computation Input: array of numbers. 2 , D. J.

The power of How fast is sorting? parallel computation Input: array of numbers. 2 , D. J.

The power of How fast is sorting? parallel computation Input: array of numbers. 2 , D. J. Bernstein Each number in 1 2 represented in binary. Thanks to: University of Illinois at Chicago Output:

786 views • 67 slides
Welcome to the January 2019 Presentation Companion Planting Agenda 2019 Presentations

Welcome to the January 2019 Presentation Companion Planting Agenda 2019 Presentations

Welcome to the January 2019 Presentation Companion Planting Agenda 2019 Presentations What is Companion Planting Benefits of Companion Planting Why Companion Planting is Effective Techniques Combinations 2019 Upcoming

385 views • 22 slides
Homogeneous Linear Systems Three Cases: Distinct Real Eigenvalues Repeated Eigenvalues

Homogeneous Linear Systems Three Cases: Distinct Real Eigenvalues Repeated Eigenvalues

Homogeneous Linear Systems Three Cases: Distinct Real Eigenvalues Repeated Eigenvalues Complex Eigenvalues MSW MTH 291 Distinct Eigenvalues General Solution-Homogeneous Systems Let 1 , 2 , . . . , n be n distinct real

264 views • 4 slides
splitSVM: Fast, Space-Efficient, non-Heuristic, Polynomial Kernel Computation for NLP

splitSVM: Fast, Space-Efficient, non-Heuristic, Polynomial Kernel Computation for NLP

splitSVM: Fast, Space-Efficient, non-Heuristic, Polynomial Kernel Computation for NLP Applications Yoav Goldberg Michael Elhadad university-logo ACL 2008, Columbus, Ohio Yoav Goldberg, Michael Elhadad splitSVM: Fast SVM Decoder

288 views • 12 slides

More recommend